Ensemble Machine Learning-Based Intrusion Detection System for Securing the Internet of Things

With the rapid growth of smart devices, IoT networks have become an essential part of our daily lives, but they're also increasingly vulnerable to cyberattacks. Traditional intrusion detection systems (IDSs) often struggle to keep up with the growing complexity of these threats, leading to delays in detection and a high rate of false alarms. To tackle this, we developed an Ensemble Intrusion Detection System (EIDS) that combines the strengths of several machine learning algorithms through a Stacking Classifier. At its core, our system uses Logistic Regression as a meta-learner and integrates Support Vector Machine (SVM), Random Forest (RF), K-Nearest Neighbor (KNN), and Gradient Boosting (GB) classifiers. By addressing data imbalance with the Synthetic Minority Oversampling Technique (SMOTE) on the NSL-KDD dataset, we ensured fair and effective training. Our approach achieves an impressive 97.62% accuracy, making it not only more reliable but also better at minimizing false positives. This solution brings us one step closer to securing IoT networks against the evolving landscape of cyber threats.